On Tuesday 29 July 2003 19:18, Ariz Jacinto wrote:
>
> hindi ba pang MTA/POP3/IMAP ito?
depends on the client. kmail 1.5 (using kde 3.1.0)
has a "pipe through" filter action. you can do that.
e.g., create a condition that hits everything (e.g.,
message size is greater than 1 byte), then set its
action to be piped through to spamc. then let
it continue processing (so that further down, filters
for the spam headers can handle the email if
it's spam).
when the email comes out of the pipe, the spamassassin
X-Spam-Status and other markup is in the email already.
evolution 1.2.2 has a "pipe message to shell command",
but it doesn't seem to do what i want (i guess it's
piping the message, but it's not putting the spamassassin
markup into the message). i can think of workarounds,
but i won't try any of that right now since i already have
a working system and i don't want to jinx it :).
i don't use any other mailers, but i'd say that support for
piping to an external program is probably not that common
yet. i don't see anything like piping in opera's M2 mailer,
for instance.
since you're on linux though, it would be simpler to do
something like this:
1. run an MTA and POP3 server on the box
2. run fetchmail to fetch email from all the POP3
accounts i have (including eudoramail, although there's
a limit to the number of times a day you can connect
to them via POP3).
3. run gotmail to fetch hotmail email
4. run yahoopops to get yahoomail
fetchmail delivers to the local postfix server. and then
i just have a .procmailrc that passes everything through
spamassassin (well, it does more than that, but you get
the point).
> can spamassassin be personalized per client?
each user can have his/her own .procmailrc. they can also
have personalized spamassassin configs (e.g., individual
weights on particular criteria, individual wordlists for the
bayesian classifier, etc). but that stuff i don't touch, usually
just going with the defaults. i doubt if users will want to play
with those. but, sure, they can have their own setups.
or you could just run spamassassin at your common mail
server and have that server be the one to run fetchmail.
that way, all that administration and email aggregation
is done at a single server and the clients don't need to run
extra services which *might* be compromised but which
certainly need watching in case they need to be upgraded
due to security issues.
tiger
--
Gerald Timothy Quimpo gquimpo*hotmail.com tiger*sni*ph
http://bopolissimus.sni.ph an xcdngl nntrstng jrnl
Public Key: "gpg --keyserver pgp.mit.edu --recv-keys 672F4C78"
Men marry because they are tired; women because they are curious.
Both are disappointed.
Oscar Wilde
--
Philippine Linux Users' Group (PLUG) Mailing List
[EMAIL PROTECTED] (#PLUG @ irc.free.net.ph)
Official Website: http://plug.linux.org.ph
Searchable Archives: http://marc.free.net.ph
.
To leave, go to http://lists.q-linux.com/mailman/listinfo/plug
.
Are you a Linux newbie? To join the newbie list, go to
http://lists.q-linux.com/mailman/listinfo/ph-linux-newbie
